So here I sit with a perfectly ripe Hass avocado and a perfectly ripe fuyu persimmon. As much as I appreciate the flavor of each separately, it seems like they would go super good together. But the pale green of the avocado next to the warm orange of the persimmon is not appetizing. So here I sit thinking about a solution to my color mismatch.

Maybe the answer is, like with two fighting kids, to separate them. If I place a pile of avocado on one side of a white plate and a pile of persimmon on the opposite side, the combination doesn’t look too bad. It’s not just that there is visual space between the piles, but that space is white. I need to add a third ingredient to my combination, and that ingredient needs to be white.

What are some white ingredients that I could add? Cooked egg whites? Wrong texture. Lightly whipped cream? Too much fat in mouthfeel. Potato? Not white enough and needs perfect cooking. Jicama? Too crunchy. Meringue cubes? Too sweet. Buttermilk? Hmmm. But in what form?

I had planned to dice both the persimmon and avocado into very small cubes. Should I make the buttermilk into similar squares? It wouldn’t be hard. Just add some gelatin or agar to the buttermilk and cast a thin sheet. When solidified, the buttermilk sheet could be cut into cubes. Would that work? After further consideration, I don’t think this is the way to go. With gelatin, the texture would start to remind me of egg whites. With agar, the corners would be very fragile. In engineering, we always round sharp corners to protect them. Can I round the corners of the buttermilk-agar cubes? I think I can.

When I attended Dave Arnold’s hydrocolloids class in January, 2010, in New York City, before lunch one day he described a technique for making spheres where a hot liquid with agar in it was dropped into a deep container of very cold oil. As the droplets floated down in the oil, they formed spheres and gelled. He was going to demonstrate the technique after lunch, but one of his assistants mistakenly threw the cold oil away during lunch. I only had the description to go on.

I filled a jar with canola oil and placed it in my refrigerator, which is set at 3 °C (38 °F). The next day I measured out 100 ml (3-1⁄3 fl oz) of low-fat, cultured buttermilk into a saucepan. I added 1 g (about 3⁄8 t) powdered agar to the buttermilk, and brought both to a boil. Damn. As the buttermilk started to boil, it began to separate. So I vigorously stirred it, and loaded a syringe with the troubled liquid. I tried various drop sizes and rates. They all seemed to work. I strained the spheres from oil, and returned it to my refrigerator. I tossed out the troubled spheres.

Day two began as a repeat of day one, except I added 2 g (1⁄2 t) tapioca starch to the saucepan along with the buttermilk and agar. This time, when the solution came to a boil, the buttermilk didn’t separate. I proceeded to make spheres the same way as on the previous day. I found that I could produce smaller spheres by holding the syringe high above the oil. I also found that I was heating the oil by adding all the hot buttermilk solution. By the time my syringe was empty, the larger spheres were no longer solidifying all the way through before reaching the bottom of the oil jar. Oh well.

I strained the spheres from the oil, and left them to drain and fully gel in my refrigerator. When I checked them the next day, all was well. I had my white separator.

When it came time to serve the dish, I diced both the avocado and persimmon into 3-mm (1⁄8-in) cubes. I drizzled a few drops of Japanese rice vinegar and a small pinch salt over the avocado to both bring out its flavor and to keep the cubes separated.

Each of the three components was carefully spooned into the serving glasses. First the avocado cubes, then the buttermilk spheres, finishing with the persimmon cubes. As each was added, I gently pressed down on the top to level the layer.

The glass was served with a small spoon, and my guests were invited to mix the contents, if they wished, before devouring. Which they did.